8.D.1 Bracketing: determination of critical substances

"Cleaning procedures for products and processes which are very similar, do not need to be individually validated. It is considered acceptable to select a representative range of similar products and processes concerned and to justify a validation programme which addresses the critical issues relating to the selected products and processes. A single validation study under consideration of the "worst case" can then be carried out which takes account of the relevant criteria. This practice is termed "Bracketing"."
(PIC/S PI 006, Section 7.3.5).

This procedure, described for the first time in the previously mentioned PIC/S directive, can drastically reduce the scope of the cleaning validation - as only a justified number of critical products (worst case products) must be considered for the validation, and not each individual product.

How is it possible to identify these critical products amongst all the products manufactured at a company? A solution for this is presented below.

There are generally three reasons why a product is assessed as critical (see figure 8.D-1):

Risk assessment - first step

The first step of the risk assessment involves assigning all products to one (or several) of these risk groups. In so doing, it may prove useful to define a limiting criterion for each group, e.g. "solubility < x" or "therapeutic dose < y". This makes it possible to pre-select critical products directly and provides a clearer overview of subsequent steps. The table in figure 8.D-2. shows how this grouping may be carried out.

In this example, products A, C, E and H must be assigned to the solubility risk group as the active pharmaceutical ingredient they contain is practically insoluble. Products B, E and F contain active pharmaceutical ingredients with a therapeutic dose of less than 100 mg and are therefore assigned to the pharmacology risk group. Products C and G contain formulation components that are difficult to remove and therefore belong to the formulation risk group. Products D and I can be excluded at this stage from the cleaning validation as they do not correspond with any of the criteria used for assignment to risk groups.

Note: the limiting criterion of 100 mg for the pharmacology risk group was selected at random. In practice, a value is recommended that is within the lower third of the available dosing range - with respect to the entire product range.

Risk assessment - second step

In the second step, the critical or worst case products in each risk group must be identified. The following selection criteria may be used to do this, for example (see figure 8.D-3):

The following example (see figure 8.D-4) shows how the worst case products within the various risk categories may be selected.

First, consider the solubility risk group. As only products containing practically insoluble active pharmaceutical ingredients are included here, content of active pharmaceutical ingredient and therapeutic dose are the only additional selection criteria used.

Product A is the worst case with respect to the content of active pharmaceutical ingredient (50 %) in this risk group, and product E is the worst case with respect to therapeutic dose (10 mg). Products C and H may be excluded at this point as they are less critical than product A in terms of active pharmaceutical ingredient content, and less critical than product E in terms of the therapeutic dose of the active pharmaceutical ingredients.

Within the risk proup "pharmacology", product F has the highest content of active pharmaceutical ingredient (50 %). Product E is the worst case in the risk group in terms of the solubility and therapeutic dose of its active pharmaceutical ingredient.

The result shows that, for bracketing, products A, E and F represent the worst case, taking into account all relevant criteria (here, active pharmaceutical ingredient content, solubility and therapeutic dose).

The cleaning validation is carried out only for these critical markers. If a cleaning procedure has been validated for the worst case products, this validation also applies for all other less critical products to be cleaned using the same procedure.

Bracketing considerably reduces the scope of the cleaning validation.

The formulation risk group represents a special case as the anticipated residues originate from excipients and not from active pharmaceutical ingredients. However as this problem occurs fairly regularly in practice, a solution should be indicated.

The formulation for product C contains a soluble dye. A coloured solution is therefore produced during cleaning. If this is not properly removed during rinsing, dried-on residues may cause undesirable coloration of the subsequent product. Though this is not a genuine active pharmaceutical ingredient cross-contamination, the quality of the subsequent product will still be adversely affected. It is therefore recommended in such instances that a cleaning validation be carried out although certain acceptance criteria and analysis aspects may be omitted. A pragmatic yet efficient means of testing this would be, for example, to wipe critical areas with a moist white cloth which means that even tiny quantities of leftover dye would be detected. This procedure corresponds with the visually clean criterion which can certainly be regarded as sufficient in this context.

The film coating of product G contains an insoluble pigment. In contrast to product C, where all production equipment may be affected by residues, validation activities for product G are restricted to the coater. The visually clean criterion may also be applied here and a suitable test method is to rub the surface with a dry cloth.

8.D.2 Matrixing: determination of equipment-specific
validation protocols

In principle, two possible approaches may be taken when implementing the cleaning validation in practice once the worst case products have been identified:

• product-specific implementation
• equipment-specific implementation

For the product-specific approach, a validation protocol is compiled for every critical product that includes all types of equipment necessary for the production process.

For the equipment-specific approach, a validation protocol is compiled that includes all critical products manufactured on this equipment.

Either approach may be beneficial, depending on the operational circumstances and scope of the cleaning validation.

However, regarding the product-specific approach, the following potential disadvantages must be considered:

• Several cleaning processes are dealt with in the validation protocol. If problems occur during the validation of one of the cleaning processes, the validation of all other cleaning processes included in the validation protocol may only be completed once these problems have been resolved, further tests have been carried out and the final validation report can be compiled.
• The validation protocol covers several types of production equipment and may - depending on the product - be extensive and confusing.
• Where changes to a specific type of production equipment and/or its cleaning process are made, all product-specific validation protocols in which this equipment is listed must be checked and revalidated, if required.

By comparison, equipment-specific implementation offers the following benefits:

• the validation protocol only deals with one type of production equipment, and therefore only one specific cleaning process. Several cleaning procedures may be validated in parallel, and independently of, one another. Problems that arise during validation of any of the procedures will not delay the completion of validation work for other procedures.
• Where changes are made to specific production equipment and/or its cleaning procedure, only one validation protocol is affected and has to be checked for possible revalidation.

Irrespective of which approach is selected, a validation matrix must initially be compiled that shows which production equipment will be used to manufacture the critical products.

The following example (see figure 8.D-5) shows what this kind of matrix can look like. It is assumed here that worst case products from the Bracketing chapter are solid dosage forms with different manufacturing processes.

The entire process flow must be represented when compiling the validation matrix, even if individual items of production equipment (in this case container type A and B) appear several times. This guarantees that the maximum number of critical products for each type of equipment will be taken into account. This table may also be used to calculate the limits later on.

Figure 8.D-5 shows that in the example chosen either

• 3 product-specific validation protocols incorporating 6-9 cleaning procedures

or

• 10 equipment-specific validation protocols incorporating 1-3 critical products

must be compiled.

The validation matrix may be used as the basis for the compilation of time schedules and capacity estimates (see chapter 8.C Cleaning validation master plan).